Approximately one-third of women with chronic pelvic pain have endometriosis. Surgical treatment of endometriosis with laparoscopic excision of implants and lysis of adhesions is often successful in reducing pain in the short term. Furthermore, several postoperative medical treatments have been shown to be efficacious in maintaining pain reduction for as long as the medication is continued. After stopping the medication, however, the level of pain tends to trend upwards towards pre-treatment levels. The desired postoperative treatment is one that is can be used for a long period of time, so as to minimize the chance of pain recurrence. Simplicity of administration and cost-effectiveness are other desirable characteristics of an ideal postoperative regimen. The only FDA approved 12-month treatment for endometriosis-associated pain is a combination of leuprolide acetate (a GnRH analog), 11.25 mg IM q 12 weeks, and norethindrone acetate, 5 mg PO daily. Use of this regimen has been constrained by its complexity and cost. As an alternative, the continuous use of oral contraceptives has been advocated as a practical, inexpensive strategy for long-term medical treatment of endometriosis-associated pain. Although such an approach is frequently used in clinical practice, there has been no clinical trial of its efficacy. The goal of this project is to compare the efficacy and cost-effectiveness of continuous oral contraceptives and leuprolide+norethindrone in the postoperative treatment of endometriosis-associated pelvic pain. Investigators at the University of Rochester School of Medicine and Harvard Medical School will recruit 194 women for randomization to one of the two treatments, each of which will continue for 48 weeks. Randomization will occur after biopsy-proven endometriosis is established. Pelvic pain and quality-of-life assessments will be obtained at regular intervals. The recruitment goal is 88 subjects per arm after a 10% drop-our rate. Change scores for these measures will be compared between the two treatments. The study has 80% power to test non-inferiority of oral contraceptives at the 5% level of significance, using a 1-point difference in the change scores for pain as the threshold. In addition, a cost-effectiveness analysis will be performed by calculating the cost per unit reduction in pain score for each treatment.